The latest fashion retail market research insight from Rebecca Harris.
In this latest piece, Rebecca Harris shares insight from Mustard’s AI in Retail consumer research into one of fashion retail’s most persistent commercial challenges: returns. This is not a tech feature or a celebration of shiny new tools. It is a margin story grounded in the latest retail market research. It looks at behavioural change, commercial impact and what fashion brands should be doing next.
According to Mustard’s AI in Retail consumer survey, 28% of 25–34-year-olds say AI has reduced the number of items they’ve had to return through better guidance and visualisation of products on their body.
This is significant. In an industry where online return rates routinely sit between 30–50%, this is not a marginal improvement. It is an early signal of structural change for fashion eCommerce. For years, fashion has tolerated returns (of online purchases) as a cost of doing business. Retailers have optimised reverse logistics rather than questioning the model itself. But now… AI is quietly dismantling that model.
AI Is Reducing Uncertainty at the Point of Decision
The 25–34 cohort matters here. This is the demographic most comfortable with digital tools and the one shaping expectation curves for the rest of the market.
When over a quarter say AI has already reduced their returns, it tells us something important for retail research teams and commercial leaders alike: AI is changing behaviour, not just experience.
Two shifts are driving this
1. Fit Intelligence Is Replacing Guesswork
Size charts were always a compromise. They are static, generic and detached from the individual. Now, retailers like ASOS are embedding AI-driven fit recommendations that learn from:
• Purchase history
• Body data
• Brand-specific cut variations
• Peer return patterns
The result is fewer speculative purchases. Less ordering multiple sizes with the intention of sending most of them back.
For fashion market research teams, this highlights a critical point. Data-led personalisation is no longer a brand enhancer. It is a cost reducer and margin protector.
2. Visualisation Has Become Personal, Not Aspirational
For decades, fashion retail has relied on aspirational imagery. But aspiration does not reduce uncertainty. What reduces uncertainty is AI-powered visualisation of items on our own bodies, from augmented reality try-ons to body-based avatars.
Brands such as Zara and Nike are experimenting with more immersive try-on experiences, while platforms like Snap Inc. are normalising AR commerce within everyday browsing behaviour. The shift is significant. Consumers are no longer imagining how something might look. They are previewing it. That confidence influences purchase decisions and return behaviour.
The Commercial Implications Are Substantial
If AI reduces returns even incrementally, then:
• Reverse logistics costs reduce
• Inventory distortion reduces
• Full-price sell-through improves
• Customer lifetime value increases
In a climate of compressed margins and rising fulfilment costs, reducing returns is one of the few levers that meaningfully impacts the bottom line without increasing price.
From a retail market research perspective, this is where the focus should sit. The question is not whether AI is interesting… it is whether it is improving commercial performance.
What Should Retailers Do Now?
First, trail new AI tools and measure the impact. Go beyond adoption metrics and track behavioural changes. Are return rates falling among users of AI-powered tools? Are specific categories benefiting more than others?
Second, integrate insight across teams. Returns data, customer research and AI performance metrics should sit in the same conversation. Too often they do not.
Third, prioritise infrastructure over theatre. The competitive advantage here will not come from a press release. It will come from embedding AI into fit, visualisation and product discovery in a way that quietly improves decision confidence.
The retailers that win will notice:
- Fewer returns
- Cleaner inventory flow
- Stronger margins
- Higher customer confidence
And over time, the impact will compound. 28% is not an end state… rather, it is an early indicator. The retailers that treat AI as core retail infrastructure will build a more resilient fashion model.
Because in fashion, the most profitable sale is the one that does not come back.
If you would like to explore what this means for your brand, or understand how AI is shaping behaviour in your category, get in touch. Our regular fashion market research and insight programmes have been helping brands move beyond technology hype and focus on solutions that drive measurable commercial impact.